Numbers 4: Intimidation vs. Interpretation

This week, our task was to interpret a set of data and  form a null hypothesis and alternate hypothesis for the data. A null hypothesis suggests there is no significant difference between two groups being compared. I liked the comparison to the grumpy old man...very stubborn.  An alternative hypothesis  for a set of data suggests there is a difference between the sets.  For this set of data, the following 2 hypothesis' are listed:

Null: There is no difference in race/ethnicity in 8th grade Reading Test scores in 2013.

Alternate: Race/ethnicity has an effect on 8th grade Reading Test scores in 2013.

After viewing Dr. Pierce's Screencast about how to perform a T-Test through Excel, I jumped right in. No more intimidation for Excel assignments for me. I used the data analysis tool to t-test White/Hispanic, Hispanic/Black, and White/Black scores on the 2013 8th Grade NAEP Reading Test. A simple process in itself was supported by the easy-to-follow screencast. Here's a look at my 3 T-Tests:

Each P value was less than .05 or 5%, which allows us to say with 95% degree of certainty that if these samples of student populations were tested again, the test would yield similar results. I am wondering if this t-test analysis proves a test's reliability? From what I understand, reliability is the sample tested will show very similar results.

This activity is relevant to teachers today because we are modeling digital age work(NETS 3a) using relevant statistics, especially when it comes to discussing Title 1 Services vs. Ethnicity/Demographics/Etc. These T-tests could also be used to collaborate with peers(NETS 3b) considering the scores between each classroom based of female/male, hours spent in instruction, etc.

Bottom Line: This Excel stuff is really growing on me.